Add multiple threshold clustering to linker

docs/api_docs/clustering.md

@@ -1,10 +1,13 @@
 ---
 tags:
   - API
-  - clustering
+  - Clustering
 ---
+
 # Documentation for `splink.clustering`
 
+Clustering at one or multiple thresholds is also available without a `linker` object:
+
 ::: splink.clustering
     handler: python
     options:

docs/api_docs/linker_clustering.md

@@ -0,0 +1,19 @@
+---
+tags:
+  - API
+  - Clustering
+---
+
+# Methods in Linker.clustering
+
+Use the result of your Splink model to group (cluster) records together. Accessed via `linker.clustering`
+
+::: splink.internals.linker_components.clustering.LinkerClustering
+    handler: python
+    filters:
+      - "!^__init__$"
+    options:
+      show_root_heading: false
+      show_root_toc: false
+      show_source: false
+      members_order: source

mkdocs.yml

@@ -140,7 +140,7 @@ nav:
       - Training: "api_docs/training.md"
       - Visualisations: "api_docs/visualisations.md"
       - Inference: "api_docs/inference.md"
-      - Clustering: "api_docs/clustering.md"
+      - Clustering: "api_docs/linker_clustering.md"
       - Evaluation: "api_docs/evaluation.md"
       - Table Management: "api_docs/table_management.md"
       - Miscellaneous functions: "api_docs/misc.md"
@@ -151,6 +151,7 @@ nav:
         - Exploratory: "api_docs/exploratory.md"
         - Blocking rule creator: "api_docs/blocking.md"
         - Blocking analysis: "api_docs/blocking_analysis.md"
+        - Clustering: "api_docs/clustering.md"
         - SplinkDataFrame: "api_docs/splink_dataframe.md"
         - EM Training Session API: "api_docs/em_training_session.md"
         - SplinkDatasets: "api_docs/datasets.md"

splink/clustering.py

@@ -1,3 +1,9 @@
-from .internals.clustering import cluster_pairwise_predictions_at_threshold
+from .internals.clustering import (
+    cluster_pairwise_predictions_at_multiple_thresholds,
+    cluster_pairwise_predictions_at_threshold,
+)
 
-__all__ = ["cluster_pairwise_predictions_at_threshold"]
+__all__ = [
+    "cluster_pairwise_predictions_at_threshold",
+    "cluster_pairwise_predictions_at_multiple_thresholds",
+]

splink/internals/clustering.py

@@ -474,7 +474,7 @@ def cluster_pairwise_predictions_at_multiple_thresholds(
         edge_id_column_name_left,
         edge_id_column_name_right,
     )
-
+    logger.info(f"--------Clustering at threshold {initial_threshold}--------")
     # First cluster at the lowest threshold
     cc = cluster_pairwise_predictions_at_threshold(
         nodes=nodes_sdf,